Product Updates: Quick fix issues by prompting your agents and follow a live tree of agents in your pentest
We have two exciting product updates to make your security coverage and workflows that much more efficient and in-depth.
What's new
Quick fix issues in Cursor, Claude, Codex or copy the prompt to paste into your agent
Escape now allows developers to prompt their agents with details regarding an issue straight from the platform. The prompt is created from an issue's context, CVSS score, asset and target URL, CVEs, root causes, its security relevance, the issue summary and its impact.
Cursor, Codex, and Claude are supported directly and you can also just copy the prompt to paste into whichever agent your team uses. To make successive fixes smoother, you can also set a default agent instead of choosing every time.
In-depth reasoning logs for all AI pentesting engagements in Cascade
You can now follow an AI pentest the way it actually runs as a live tree of agents in the new AI reasoning logs tab. During a Cascade assessment, Escape spawns an orchestrator who oversees the engagement and specialized agents for different vulnerabilities. The reasoning tab now maps this hierarchy so you can expand branches, click any agent, and read its reasoning, tool calls, screenshots, and errors. This also updates in real time so you can see what is happening and why, exactly as it happens.
Why we implemented them
These are the key benefits of the new features:
- Faster triage: Prompting your agents directly through the Escape platform means you can save precious time crafting a prompt yourselves and get to what actually matters - fixes. Similarly, the new reasoning logs feature shows an issue count badge, so reviewers can jump straight to the agents that found something meaningful.
- Trust and transparency: Now security teams can see why AI agents in Cascade chose a certain path, how it adapted when something failed and how multi-step exploits were changed to get that much deeper of an understanding of your engagement.
- Smoother developer handoff: Agent-ready prompts close the gap between a CVSS score and raw payload into something a developer can act on, so findings move into fix cycles faster with less context lost in handoff.
With these new updates you should be able to triage faster and gain a deeper insight into your security posture. Get a personalized demo to discover exactly how Escape can optimize your security.
💡 Check out more relevant articles below:
How Escape helped Bridgetech Group replace pentests with automated evidence
How Escape AI Pentesting Exploited SSRF in LiteLLM
Burp Suite Alternatives: The Complete 2026 comparison between Escape and Burp Suite